Received: 5 November 2018 | Revised: 5 February 2019 | Accepted: 26 February 2019 DOI: 10.1111/fme.12346

ORIGINAL ARTICLE

The hidden value of artisanal fisheries in

Steven Canty1,2 | Manuela Funes3 | Stephen Box4 | Kyrstn Zylich5 | Brittany Derrick5 | Esther Divovich5 | Alasdair Lindop5 | Daniel Pauly5 | Dirk Zeller6

1Smithsonian Marine Station, Fort Pierce, Florida Abstract 2Manchester Metropolitan University, Declining fisheries catches are a global trend, with management failing to keep pace Manchester, UK with growth in fishing effort and technological advances. The economic value of 3CESIMAR‐CONICET, Puerto Madryn, Honduras’ catches was estimated within the industrial and artisanal sectors. Catches 4Rare Inc., Arlington, Virginia were found to be 2.9 times greater than the official statistics between 1950 and 5Sea Around Us, University of British 2015. The merging of industrial and artisanal catch data masked the decline in indus‐ Columbia, Vancouver, British Columbia, trial catches and hid the strong growth of artisanal fisheries. In 1996, annual artisanal Canada 6Sea Around Us–Indian , School of fisheries landed catches surpassed the industrial fishery sector, and in 2000, the an‐ Biological Sciences, University of Western nual net value of artisanal fisheries eclipsed the value of the industrial fisheries. , Crawley, Western Australia, Australia These data highlight the importance of artisanal fisheries in Honduras and challenge the long‐held belief that the industrial sector contributes more to the national Correspondence Steven Canty, Smithsonian Marine Station, economy. The global paucity of fisheries data highlights the need for comprehensive Fort Pierce, FL. strategies to collect more detailed and accurate fisheries data. Email: [email protected]

KEYWORDS Funding information Summit Foundation; Marisla Foundation; artisanal fisheries, catch reconstruction, fisheries data, industrial fisheries, IUU fishing, marine Oak Foundation; Paul M. Angell Family fisheries, small‐scale fisheries, subsistence fisheries Foundation; Consejo Nacional de Investigaciones Científicas y Técnicas; David and Lucile Packard Foundation

1 | INTRODUCTION stock management and the potential for cross‐border fishing ac‐ tivities (Perez, 2009). The Republic of Honduras is located on the Fishing has been an important economic sector in Honduras for Isthmus, with coastlines on both the Atlantic and Pacific more than 100 years (MacKenzie & Stehlik, 1996), contributing 5% (Figure 1). In accordance with the United Nations Convention of the country's gross domestic product with an average value of on the Law of the , Honduras claimed its exclusive economic US $385 million per year (Beltrán Turriago, 2011). Fishers employ a zones (EEZs) in both the Atlantic (within FAO statistical Area 31) diverse range of gears to exploit and riverine systems, coral and the Pacific (within FAO statistical Area 77). The north shore reefs, other near‐shore habitats (e.g. beds), extensive off‐ of Honduras, located within the basin of the Atlantic shore banks and pelagic waters (Box & Canty, 2011; Soto, 2012). The Ocean, is the longer of the two coastlines and is bordered by the main fisheries on the Caribbean coast are for Caribbean spiny lob‐ EEZs of , , Mexico, Cuba, the Cayman Islands, ster, Panulirus argus (Latreille), and queen conch, Lobatus gigas (L.), Jamaica and . The Pacific coastline is much smaller while the main fishery on the Pacific coast targets western white and is exclusively within the Gulf of Fonseca, fully enclosed by shrimp, Litopenaeus occidentalis (Streets) (FAO 2002). the EEZs of and Nicaragua (Figure 1). The sharing of Artisanal and subsistence fisheries have been present in EEZ borders with numerous countries, particularly within the Honduras at least since the Mayan era, however, the large expan‐ Honduran Caribbean, has implications for transboundary fish sion of artisanal fisheries across the Caribbean and Pacific coasts

© 2019 John Wiley & Sons Ltd. This article has been contributed to by US Government employees and their work is in the public domain in the USA.

Fish Manag Ecol. 2019;1–11. wileyonlinelibrary.com/journal/fme | 1 2 | CANTY et al. is a relatively recent phenomenon, commencing in the 1970s most fundamental of all fisheries data. Comprehensive and accurate (MacKenzie & Stehlik, 1996). Artisanal fisheries provide an essential records of fisheries’ catches are important to allow monitoring of fish‐ source of nutrition and employment to coastal communities, espe‐ eries’ trends over time, with the effect of fisheries’ regulations to be cially in remote coastal areas where there are few other economic observed, and subsequently adjusted if required (Belhabib, Koutob, alternatives. Due to a lack of processing infrastructure, catches from Sall, Lam & Pauly, 2014). However, while technological advances in the Honduran Pacific coast supply only national markets (Box & fishing vessels and gears are evident, systems to record catch data Bonilla, 2009), while artisanal fisheries on the Caribbean coast sup‐ and monitor fisheries have not kept pace. There is a chronic need ply both national and international markets. In addition, the shallow for improvements in data collection and the incorporation of these hydrology precludes the deployment of larger boats within the Gulf data into fisheries management (Pauly et al., 2002). The status of of Fonseca, which restricts the Honduran industrial fishery to the the world's fisheries may be worse than currently perceived, given Atlantic coast (Soto, 2012). that a large fraction of catches has been missing from national fish‐ The Honduran national industrial fishery began in the late eries catch estimates in virtually every country of the world (Pauly & 1950s (FAO, 2002). Previously, only foreign industrial fleets, mainly Zeller, 2016a,2016b). Crucially, the pattern of missing data changed from the , were fishing and landing in Honduras. As over time as an inadvertent by‐product of well‐intentioned efforts to international fleets began to leave Honduran waters, the national improve data collection systems, resulting in a time series bias now industrial fishing fleet started to develop. An additional trigger known as “presentist bias” (Zeller & Pauly, 2018). Thus, there is a con‐ was the collapse of the United States industrial conch fishery in certed global effort led by the Sea Around Us initiative to reconstruct 1975, which enabled the Honduran industrial fishery to start sup‐ national, and by extension, regional and global fishing statistics that plying the United States market. Currently, 90%–95% of indus‐ add comprehensive estimates for all unreported catches to officially trial marine catches are exported, primarily to the United States reported landing data to derive a better and more comprehensive (Espinoza, 2007). Recreational fisheries are practised across the understanding of fisheries catches over time (Funes et al., 2015; Honduran shore in the Caribbean, but sport fishing operations are Pauly & Zeller, 2016a,2016b; Zeller, Booth, Davis & Pauly, 2007; primarily located in the Bay Islands, where the majority of inter‐ Zeller, Harper, Zylich & Pauly, 2015; Zylich et al., 2014). national tourism is concentrated, receiving over 700,000 tourists The starting point for reconstructions is the official reported per year (INE, 2018). landing data provided by national agencies to the Food and Signs of overexploitation have been documented in Honduran Agriculture Organization of the United Nations (FAO); these data fisheries. The conch fishery closed in 2003 due to a trade embargo are subsequently compared with the formal and grey literature, and placed on conch exports by the Convention on International Trade inferences on additional, previously unreported catches are vali‐ in Endangered Species of Wild Fauna and Flora (CITES, 2003). The dated with local experts (Zeller et al., 2016), Funes et al. (2015) and Nassau grouper, Epinephelus striatus (Bloch), population collapsed from reconstructed marine fisheries catch data for the Republic of in 2004 (Box & Bonilla, 2009). According to Honduran fishers, the Honduras, by deriving the best time series estimates of unreported decline and collapse of these fisheries was a consequence of ir‐ catches for 1950–2010, from both the Atlantic and Pacific coastlines responsible fishing practices and habitat degradation via destruc‐ to complement reported data. The present study revised, improved tive fishing gears (Korda, Hills & Gray, 2008), which mirrors trends and extended the analysis performed in Funes et al. (2015), by up‐ in Jamaica (Hughes, 1994) and the Caribbean as a whole (Gobert dating the time series to 2015, rectifying all estimates of fishing et al., 2005). categories, completing the lacking fishing categories and enhancing Overfishing is a great challenge facing governments and the sci‐ the sources of data. Once the corrected reconstructed catch was entific community (FAO, 2016). As a minimum requirement, effective available, the aim was to describe and compare statistically the catch management requires reliable data, with catch data representing the trends over the years, and to evaluate the economies of the fisheries

(a) (b)

FIGURE 1 Maps of the two parts of the Honduran exclusive economic zone (EEZ). (a): Honduras’ EEZ in the Caribbean covers 218,000 km2, of which 60,300 km2 is shelf, that is, less than 200 m deep. (b): Honduras’ EEZ in the Pacific is small (747 km2) and shallow, and consist only of the inner Gulf of Fonseca, shared between El Salvador in the North, Honduras, and Nicaragua in the South CANTY et al. | 3 sectors in Honduras and in the context of the Mesoamerican reef the national economy. Here, illegal catches taken by the Honduran countries (Mexico, Guatemala and Belize). fishing fleet from neighbouring EEZs of Nicaragua and Jamaica were estimated. The Honduran Grand Banks are situated in the far eastern part of the EEZ, close to the Nicaraguan border, extending 2 | METHODOLOGY north‐east to the Jamaican border, and the Honduran fleet enters into Nicaraguan and Jamaican EEZ waters illegally. To consider this 2.1 | Catch reconstructions and remain conservative, illegal catch was estimated as 15% of the Honduran catch reconstructions were conducted using the method total estimated industrial lobster catch, as a constant percentage of Zeller et al. (2007, 2016), following the principles described in over time. Estimations on foreign fishing vessels illegally entering Pauly (1998). Only marine wild capture fisheries were addressed; to Honduran EEZ were not included in the analysis; however, ille‐ therefore, aquaculture production and freshwater catches are not gal fishing within the Honduran EEZ was assumed to be equivalent included in the estimates, nor are catch records of marine mammal, to that of the Honduran fleet in other EEZs, and therefore, the il‐ turtles, worms, seaweed or algae. For a full list of the categories used legal capture of the Honduran industrial fleet was included in the in this reconstruction, see Supporting Information (Table S1). estimations. Industrial, artisanal, subsistence and recreational fisheries for fishing areas 31 and 77 (Figure 1) were estimated separately. 2.2.3 | Unreported industrial catch Artisanal fishing was defined by the Honduran Fisheries Law of 1959, as “fishing within three nautical miles from the shore using Undeclared discards are part of the industrial unreported catch in boats with a capacity of three tonnes or less and employing basic Honduras. Discards are absent or minimal in the conch and lobster fishing equipment.” All commercial fishing activities encompassed by fisheries, as conch are hand‐collected and lobster are either hand‐ this definition were considered artisanal. Commercial fisheries with collected or caught in traps. The capture of finfish is mainly from vessels of greater capacity and fishing at greater distances were con‐ vertical long lines, where low‐valued bycatch is usually consumed sidered industrial. Subsistence fisheries were defined when landed and accounted for in the subsistence fishery category (see below). catch was for personal consumption and no commercial transaction The major non‐selective industrial fishery component is the bottom‐ is associated with it. Recreational fisheries were defined as sport trawl shrimp fishery. Unregulated shrimp discards were estimated fishing; whereby, individuals may pay for a fishing excursion, but re‐ following Davies, Cripps, Nickson, and Porter (2009), who suggested tained landed catch was not sold. a bycatch rate of 78%, where no bycatch was landed. This percent‐ age was applied to FAO reported shrimp landings for 1950–2015 to estimate what was taken from the sea to be discarded later. Any 2.2 | The (FAO Area 31) bycatch in the industrial finfish fishery was assumed to be either re‐ corded in the official records or used for personal consumption and 2.2.1 | Reported industrial catch therefore estimated under subsistence. Industrial fishing in Honduras is limited to Area 31 (Caribbean) with four fleets licensed to exploit four resource types: Caribbean spiny 2.2.4 | Reported artisanal catch lobster, queen conch, shrimp and finfish (primarily grouper and snap‐ per). The industrial sector is diverse, fishing vessel lengths vary from The large extent of the Caribbean coastline provides suitable fishing 4 to 78 m (13–258 feet), crews range from 6 to 85 people, and the grounds across the Honduran mainland and island archipelagos, with duration of fishing excursions vary from 10 to 90 days, depending on more than 135 different fishing communities (Stephen Box, personal the fishery (CITES, 2003). National industrial fisheries first appeared observation) with over 7,000 registered artisanal fishers included in in Honduras during the late 1950s (FAO, 2002), at which time only the DIGEPESCA fisher registration system. The geographical disper‐ limited catch data were reported. Given that industrial catches were sion and remoteness of many of these communities have made it always landed at major ports, all catches from 1950 to 1989 reported difficult to collect catch data from these fisheries, resulting in limited to FAO were considered industrial. In 1990, the governmental body available data. DIGEPESCA was created and began to collate fisheries data; there‐ From 2001, FAO records included some artisanal fishing ac‐ fore, for the period of 1990–2015 industrial landed catch data were tivity (Diana Vasquez, Centro de Estudios Marinos, personal obtained directly from DIGEPESCA records. communication). It was assumed that the remainder of FAO catch unaccounted in DIGEPESCA data was artisanal catch for the pe‐ riod of 2001–2015, where FAO landings exceeded DIGEPESCA re‐ 2.2.2 | Illegal industrial catch ports. Since artisanal fishing in this was not reported until Illegal catch is any catch that is acquired or removed from the EEZ recently, the FAO landing data for the early time periods were as‐ of a country without the consent of that country. In instances where sumed to be industrial fisheries. As a conservative estimate, the these landed catches are processed through Honduran packag‐ reporting of artisanal fisheries to the FAO was considered to im‐ ing plants, they become included in official data and contribute to prove linearly from 0% of catch reported to the FAO in 1989 to 4 | CANTY et al. the calculated percentage of FAO catch estimated to be artisanal on the Atlantic coast, which also matched well with grey literature in 2001 (i.e. 74% of reported data). estimates (Box & Canty, 2011). The artisanal fisher population was assumed to change over time as a fixed proportion (0.00177%) of the national population of Honduras from 1950 to 2015 (World Bank, 2.2.5 | Illegal artisanal catch 2017). The per‐fisher level of subsistence catch was taken from a Transboundary fishing by artisanal fishers is a problem with study conducted in the neighbouring country of Guatemala (Trujillo, Honduran boats making excursions into Belizean territorial waters Cisneros‐Montemayor, Harper, Zylich & Zeller, 2012), which esti‐ and Guatemalans fishers making excursions into Honduran (and mated a take‐home subsistence portion of 70 kg/fisher/year. This Belizean) waters (Perez, 2009). To incorporate this and remain con‐ estimate is conservative considering that fishers share the catch servative, illegal artisanal catch was estimated as 15% of the total with their family. estimated artisanal catch, as a constant percentage over time. These In addition, the common Caribbean donax, Donax denticulatus L., illegal catches were identified and labelled as Honduras fishing in is gathered by hand, usually by local women and children (MacKenzie Belizean waters. The catches of foreign artisanal fleets fishing ille‐ & Stehlik, 1996). The donax catch is intended only for subsistence gally in Honduran waters were not estimated. As per the industrial and not sold at markets (MacKenzie & Stehlik, 1996). As a minimum, fishery, illegal fishing of others in Honduran waters was assumed an estimated average of 5 kg of donax per year per artisanal fisher's to be equivalent to the illegal transboundary fishing activity of the household was used in our calculations. Therefore, a total subsis‐ Honduran artisanal fleet and therefore to include illegal landed tence catch of 75 kg/fisher/year was estimated for the Honduran catch. Caribbean region. This is likely a very conservative estimate.

2.2.6 | Unreported artisanal catch 2.2.8 | Recreational fisheries catch

Despite artisanal fisheries existing in Honduras since the pre‐ Recreational fisheries were categorised as the capture and non‐re‐ Columbian period, no data were available for the artisanal fishery for lease of sport fish species (e.g. mahi mahi, Coryphaena hippurus (L.)). the period 1950–2000. Reporting of artisanal fisheries has occurred To estimate catch from this sector, Honduran sport fishing compa‐ since 2001 primarily from the cities of La Ceiba, Puerto Cortex and nies using online searches were made via Google and TripAdvisor Tela along the Caribbean coastline. However, no catch data have with combinations of the following search terms, “sport,” “fish*,” been collected from many of the smaller fishing communities in “recreation*,” “trip*,” “excursion*,” “tourism,” “activities,” “Honduras,” more remote location along the Honduran north shore, despite their “Bay Islands,” “Roatan,” “Utila,” “Guanaja,” “Cayos Cochinos” (*rep‐ considerable fishing effort (Stephen Box, unpublished data). In ad‐ resents derivatives of a search term, e.g. fish, fishing). Secondly, fol‐ dition, large inconsistencies were found between the FAO data set lowing Belhabib et al. (2016), YouTube videos were identified using of 2011 and other sources of information. For example, differences names of sport fishing companies collated in the previous search. in reported catch vary in magnitudes of 4.8–10 times, principally From each video, the number of fish caught and their species were for lobster and finfish catch (e.g. FAO, 2002; Heyman & Granados‐ identified and recorded. The weight of landed catch was sometimes Dieseldorff, 2012). To remain conservative, estimates of unreported reported by the fisher; where this did not occur, fish length and sub‐ artisanal catch assumed that reported artisanal catch reflected 50% sequently their weight were estimated using species‐specific param‐ of total catch from artisanal fisheries from 2001 to 2015. Total re‐ eters to transform length to weight from www.fishbase.org (Froese constructed artisanal catch from 2001 to 2015 was used to calculate & Pauly, 2018). The number of annual trips of each company was an average catch rate per artisanal fisher: annual tonnage of the re‐ estimated through tallying the number of reviews and Instagram ported artisanal data (2001–2015) divided by the number of fishers posts each company received per year; these were cross‐refer‐ for that time period, which led to an overall catch rate (i.e. tonnes of enced to prevent duplication of fishing excursions. The number of fish per fisher per year). In the period of 1950–2000, reconstructed reviews ranged from 16 to 18 for the most popular companies over artisanal annual catch was estimated by multiplying the catch rate the last 3 years, and the number of Instagram posts was around 150. by the number of artisanal fishers estimated for each year (FAO, Recreational tourism fishing was assumed to have started officially 2002); see methodology2 of artisanal fisher's estimation in subsist‐ in 1985, based on information gained from the sport fishing com‐ ence section. pany websites. The number of companies that offered fishing excur‐ sions was estimated to have grown linearly from 1 to 32 between 1985 and 2012, by which time all companies had been established. 2.2.7 | Subsistence fisheries The number of fishing excursions per company was set to vary be‐ Fishers from the artisanal sector may set aside a portion of their tween 20 and 33 following trends in the number of annual visitors catch for personal consumption (MacKenzie & Stehlik, 1996); the ar‐ to Honduras (INE, 2018), as validated by local consultants (Mayra tisanal fishery‐derived subsistence catch was estimated by multiply‐ Nuñez, Centro de Estudios Marinos, personal communication). ing the number of artisanal fishers by an amount of fish per capita for In addition, since 1999 an international annual fishing tourna‐ home consumption. FAO (2002) documented a total of 9,132 fishers ment is held on Roatán, targeting mainly billfishes (Istiophoridae). CANTY et al. | 5

The amount of fish taken during fishing tournaments was esti‐ was assumed to be part of the catch of subsistence fisheries. mated. Lists of the capture, discriminated by participant, in each Furthermore, as artisanal fisheries within this area employ tram‐ category, were available for past editions on the tournament's of‐ mel and cast nets rather than trawls (Box & Bonilla, 2009; Heyman ficial Facebook page. Total catch was estimated by approximating & Granados‐Dieseldorff, 2012; MacKenzie & Stehlik, 1996), dis‐ each species’ weight (available on www.fishbase.org) and consider‐ carded bycatch is minimal. ing all the participants and days of tournament. Retained catches decreased after 2009 when the main target species started to be 2.3.4 | Subsistence fisheries released (i.e. catch and release), but fishers still kept some bycatch species for example mahi mahi or wahoo, Acanthocybium solandri Subsistence fisheries catches were calculated following the pro‐ (Cuvier),. In 2011, captures from tournaments increased as the island cedures described for the Caribbean, except that the majority of of Guanaja started to host an annual sports fishing event. subsistence catch was based on the incidental fish bycatch from the shrimp fishery (see above). In this case, the anchor point for the number of artisanal fishers was 1,600 in 2004 (Soto, 2012). 2.3 | Gulf of Fonseca (FAO Area 77)

2.3.1 | Reported artisanal catch 2.3.5 | Recreational fisheries

An industrial fishery sector as defined by the Honduran Fisheries No records of recreational fishing in the Gulf of Fonseca could be Law of 1959 is not present in the Gulf of Fonseca; all data re‐ found, and local experts had no knowledge of any commercial en‐ ported by FAO are considered to be from the artisanal fisheries. terprises engaged in recreational fisheries. It was assumed that all Any Honduran‐flagged offshore vessels were not considered in this landed catch would be consumed and therefore considered any do‐ study, as they fish outside Honduran waters (Le Manach et al., 2016; mestic recreational fishing that may occur to count as subsistence Schiller, 2014). fishing, whose catch was estimated above.

2.3.2 | Illegal artisanal catch 2.4 | Fisheries economics

Transboundary fishing activities are known to occur in Pacific wa‐ Economic data (i.e. landed value) for the fisheries of Honduras, ters; fishers from Honduras make fishing excursions into the wa‐ Mexico, Belize and Guatemala were provided by the Sea Around ters of El Salvador and Nicaragua. A conservative estimate of illegal Us. The landed value of the catch is based on the ex‐vessel price catches was thought to be 5% of the total estimated artisanal finfish data which represent the prices in real 2010 US$ paid to fishers catch, a constant percentage over time. Illegal transboundary fishing at the first point of sale, combined with catch volumes, represent activity of El Salvadorian and Nicaraguan fishers within Honduran the landed value of the catch (Sumaila, Marsden, Watson & Pauly, territorial waters was assumed to be equal to the illegal fishing ac‐ 2007; Swartz, Sumaila & Watson, 2013; Tai, Cashion, Lam, Swartz tivity of the Honduran artisanal fishers, and all landed catch by the & Sumaila, 2017). The landed value data for the four countries that Honduran artisanal fleet was included within current estimates. make up the Mesoamerican reef ecoregion (Honduras, Mexico, Guatemala, and Belize) were examined, and reconstructed landed catch data—discards not included—were downloaded from the Sea 2.3.3 | Unreported artisanal catch Around Us data portal (http://www.seaaroundus.org/data/#/search). Several reports on fishing activities within the Gulf of Fonseca (Box & Bonilla, 2009; Soto, 2012) presented significantly larger catches 2.5 | Statistical analysis than reported in the FAO 2011 data set. These reports consist of annual assessments in 2004 and 2008. For each major target group Reported and unreported data from both the Atlantic and Pacific (crabs, clams, finfish and shrimp), there were differences between coasts were combined to generate a total Honduran reconstructed the catches in the assessments and the FAO data set for the same catch time series for 1950–2015. Trends of catches tonnages and years. An unreported catch ratio of the difference between the as‐ trends of reconstructed economic values (landed value of catches) sessments and the FAO data set for each of the major target groups over time, from reported and unreported catches of the industrial was estimated and applied as a fixed ratio of the reported catches and artisanal sector, were analysed using segmented linear regres‐ through 2015. Although reporting within the artisanal fishery has sions, in the R package SEGMENTED (Muggeo, 2008), and changes improved, it remains low and fluctuates year to year. To minimise in the slope were evaluated applying Davies tests in the same pack‐ any bias, a constant percentage differential between reported and age. Segmented regression models determine regression breakpoint unreported catch was applied throughout the period of 1950–2003. years that indicate a significant change in trend over time and seg‐ Soto (2012) suggested that there are no significant discards mented line slopes. Prior to analysis, all data were evaluated for nor‐ in the shrimp fishery, because the majority of the fish caught mality, and all data were identified to be within the boundaries of as bycatch is retained and consumed. Therefore, any bycatch normal distributions. 6 | CANTY et al.

FIGURE 2 Honduran catch reconstructions for the period of 1950‐2015. (a) Total combined catch from Areas 31 and 77 (Note: official reported catch black line); (b) catch reconstructions for Area 31; and (c) catch reconstructions for Area 77 (note different y‐axis scale)

3 | RESULTS 0.01% of the total catch, while by 2015 they comprised over 20% of the total reconstructed catch. No official records were identified for The total reconstructed catches as estimated here for 1950–2015 this area in 1950, while the reconstruction suggested catches of 23 were 2.9 times greater than the data reported to FAO by Honduras t in 1950 (Figure 2c). By 2015, official records reported catches of (Figure 2a). Official records reported total landed catches of ap‐ 2,000 t, while this reconstruction estimated catches of nearly 6,000 proximately 500 t in 1950 increasing to 11,079 t by 2015. This com‐ t (Figure 2c). pares to the present reconstructions that estimated total catches of Differences in trends between reported and reconstructed around 5,000 t in 1950 and 32,000 t in 2015 (Figure 2a). Catches catch time series were observed (Figure 3). Reported total catches from the Caribbean Sea in the (FAO Area 31) com‐ suggested a continuous period of growth from 1950 to 2000, after prised the greatest contribution to total Honduran catches, with which catches began to decline (Figure 3a, Table 1). These re‐ approximately 80% in 2015. For the Caribbean side (FAO Area 31), construction estimates followed a similar trend, but the period of official catch records reported approximately 500 t in 1950, which growth is shorter, 1950–1986, after which time catches declined increased to just under 9,000 t in 2015, while the catch reconstruc‐ (Figure 3a, Table 1). In the industrial fishery, reported and recon‐ tion estimated 5,000 t in 1950, which increased to over 26,000 t structed estimates followed the same trend; a period of growth in 2015 (Figure 2b). Catches from the (FAO Area 77) from 1950 to 1986 was observed, followed by a significant decline have increased in importance: in 1950, they accounted for less than (Figure 3b, Table 1). However, reconstruction estimates suggested a CANTY et al. | 7

FIGURE 3 Reported catches (grey) and total reconstructed catch estimates (black) for total catches of (a) all Honduran fisheries, (b) industrial fisheries and (c) artisanal fisheries (note different y‐axis scale). The regression trend lines (dashed lines) represent periods of catch decline or increase (identified as a breakpoint by segmented regression analyses)

much stronger decline in industrial catches (which include discards) Historically, the industrial fishery was of greatest value; however, after 1986 (Figure 3b, Table 1). While unreported industrial landings as catches declined in this fishery and increased in the artisanal fish‐ drove much of this difference, unreported discards from this sector ery, the artisanal fishery increased in value and surpassed the landed also contributed (Figure 2a). Reconstructed and reported estimates value of the industrial fishery in 2000. The industrial fishery peaked for the artisanal fisheries followed matching trends, with a period of in landed value in 1987, at US$ 59 million, and declined to just under relatively gradual growth followed by a period of more rapid growth, US$ 13 million by 2015 (Figure 4a). The Caribbean (Area 31) artis‐ starting in the early and late 1970s, respectively (Figure 3c, Table 1). anal fisheries from the same period showed an increasing value in The growth of artisanal fisheries is associated with an increased con‐ the fishery since 1950; in 1996, a significant increase in the value tribution of this sector to total landed catches of the country, and of the fishery was observed with a peak value of US$ 35 million in by 2015, artisanal catches accounted for approximately 61% (nearly 2003 (Figure 4a). The pattern of artisanal fisheries being of greater 20,000 t) of the total catch (Figure 2a). economic value was found in all four countries of the Mesoamerican 8 | CANTY et al.

TABLE 1 Results of segmented regressions and Davies tests for total, industrial and artisanal reconstructed and reported catches of Honduras for the time period 1950–2015, and for reconstructed economic landed value (2010 US$) of the industrial and artisanal fisheries

Segmented regressions

Catch sector Landed catch time series Breakpoint (year) Line Slope (t./year) Davies test

Total Reconstructed – 1 951.8 – 1986 2 −407.8 0.0001 Reported – 1 284.5 – 2000 2 −276.4 0.0001 Industrial Reconstructed – 1 706.3 – 1986 2 −909.2 0.0001 Reported – 1 226.6 – 1986 2 −261.9 0.0001 Artisanal Reconstructed – 1 105.4 – 1972 2 456.5 0.04 Reported – 1 4.9 – 1979 2 261.8 0.0002

Segmented regression

Catch sector Landed value time series Breakpoint (year) Line Slope (US$million/year) Davies test

Industrial Reconstructed – 1 1.2 – 1986 2 −2.4 0.0001 Artisanal Reconstructed – 1 0.2 – 1996 2 0.8 0.0001

reef region (Figure 4). In Mexico and Belize, artisanal fisheries have high‐value species associated with the industrial fishery, for example been of greater value since 1950, but this gap has only increased spiny lobster; in contrast to the lower priced finfish, which contrib‐ over time (Figure 4b,c). In Guatemala, as in Honduras, artisanal fish‐ ute a greater proportion of artisanal fishery catches. A similar trend eries became more valuable around 2001 (Figure 4d). has been observed in Guatemala, while artisanal fisheries have been more economically valuable than industrial catches in Mexico and Belize since 1950. These economic data highlight the importance of 4 | DISCUSSION artisanal fisheries within the Mesoamerican reef ecoregion, which supports over 2 million coastal people (Kramer & Kramer, 2002), and The reconstructed catches for the Honduran fisheries in the Atlantic highlight the need for focused management within this important (FAO Area 31) and Pacific (FAO Area 77) EEZs were 2.9 times fishing sector. greater than the catches (landings only) reported by FAO on behalf The shift in sector emphasis from industrial to artisanal fisheries of Honduras. The reconstruction illustrated that contrary to the of‐ has important implications for fisheries management, as artisanal ficial reported data, which implied a period of continuous growth fisheries have a greater potential for sustainable use of coastal re‐ from 1950 to 2000, total reconstructed catches suggested that the sources (Pauly, 2006), and to ensure the resilience of coastal commu‐ Honduran catches have actually been in decline since 1986. This de‐ nities through food security (Golden et al., 2016) and employment cline, driven by strongly declining industrial catches, was masked by (Beltrán Turriago, 2013). These findings highlight the critical impor‐ significant increases in artisanal catches starting in the early 1970s. tance of disaggregating data to fisheries sectors to identify import‐ The growth in artisanal catches meant that by 1996, catches of the ant trends and patterns within a country's fishery (Pauly & Zeller, artisanal sector exceeded the catches of the industrial fishery and 2016a) and for investing in data collection systems for artisanal fish‐ by 2015 accounted for approximately 61% of the total reconstructed eries (Pauly & Charles, 2015). The majority of artisanal fisheries in catches. Associated with the change in the dominance of landed Honduras are within the informal sector and therefore no official catch from the industrial to the artisanal sector is a distinct shift in records of catches or the associated value exist, which proliferates economic importance since 2000, with the artisanal fishery worth the underestimation of their importance to coastal communities and more annually than the industrial fishery. There is lag of approxi‐ their economies. mately 4 years between landed catch and the economic value of the The reported data for Honduras suggested mistakenly that artisanal fishery surpassing the industrial fishery. This is due to the the country's fisheries were growing until 2000, whereas the CANTY et al. | 9

FIGURE 4 Mesoamerican reef ecoregion fishery valuations, based on reconstructed landed data (discards are not included) from the Caribbean Sea fisheries of (a) Honduras, (b) Mexico, (c) Belize and (d) Guatemala (note different y‐axis scale). Landed catch values are of reconstructed catches, and data for all countries were downloaded from the Sea Around Us online database. The regression trend lines (dashed lines) represent periods of catch decline or increase (identified as a breakpoint by segmented regression analyses) in the Honduran fishery reconstruction illustrated that the growth period ended much ear‐ with reconstructed data being higher than reported data, 1.4 times lier, in 1986, followed by a much stronger decline in catches. The for Panama (Harper, Guzman, Zylich & Zeller, 2014), 2.6 times for disparity in trends between these two data sets has important im‐ Costa Rica (Trujillo et al., 2012), 3.4 times for Nicaragua (Haas, plications for fisheries management and data collection from all Harper, Zylich & Zeller, 2015) and 3.5 times for Belize (Zeller, Graham fisheries sectors. The results highlight that the aggregation of data & Harper, 2011), and about 1.5 globally (Pauly & Zeller, 2016a,2016b). from different fisheries sectors can mask important changes within Additionally, artisanal fisheries are of greatest economic importance a country's fishery (Pauly & Charles, 2015). By disaggregating the in the Caribbean fisheries of the four countries of the Mesoamerican industrial and artisanal sectors, it was possible to show how the reef ecoregion, Mexico, Belize, Guatemala and Honduras. Interannual industrial fisheries have been in decline since 1986, while catches variability on the reported and reconstructed catches trend is com‐ from artisanal fisheries, from both the Atlantic and Pacific Oceans, monly found in all catch reconstruction analysis (e.g. Trujillo et al., have increased. The combined catches from the Atlantic and Pacific 2012; Harper et al., 2014; Haas et al., 2015). This could be a phenom‐ are greater than the industrial fishery landings. Although a lack of enon based on the natural species fluctuations or a result of changing consistent fishing effort data prevented determination of whether fishing pressure due to externalities such as fuel prices. the decline in industrial catches was related to a shift in fishing ef‐ While the estimates of catches improved in terms of tonnages, fort from the industrial sector to the artisanal sector, it is likely that the taxonomic resolution of these reconstructions is low, and for the artisanal fishery has undergone independent growth in parallel, effective management, the collection of data with greater taxo‐ rather than in response to a decline in the industrial fleet. This has nomic resolution is required. Greater investment needs to be made been mirrored in the economic growth in the artisanal sector and in collecting such improved data, which should also include fishing provides further evidence for the need of comprehensive manage‐ effort data across all fisheries sectors. There have been important ment of artisanal fisheries to maintain food security and livelihoods advances to address the complex task of collecting fisheries data in coastal communities. from widely dispersed artisanal fisheries, which are characterised The general results of this catch reconstruction for Honduras are by geographical remoteness, the diversity of supply‐chain partici‐ comparable to other reconstructions for Central American countries, pants (commercial fish buyers, markets and restaurants) and fishing 10 | CANTY et al. gears used. Specifically, in Honduras and Belize a novel freeware Beltrán Turriago, C. S. (2011). Value chain analysis of international fish application, OurFish (https://ourfish.org/), is being implemented trade and food security in the Republic of Honduras (Fisheries and Aquaculture Department Products, Trade and Marketing (FIPM)) (p. that records transactions between fishers and fish buyers, including 73). Washington D.C.: FAO. Retrieved from http://www.fao.org/ restaurants, using widely available and extensively used cell phone valuechaininsmallscalefis heries/participatingcountries/honduras1/ and Wi‐fi technology. This application is connected to national fisher en/ licensing databases, which links each transaction to individual fish‐ Beltrán Turriago, C. E. (2013). Tendencia de la actividad pesquera y acuícola en Honduras. (Contribución de la pesca y la acuicultura a la ers, thus allowing for fisheries data at the individual, community, seguridad alimentaria) (pp. 56–67). Tegucigalpa, Honduras. region and national level to be collated and used for management. Box, S. J., & Bonilla, S. (2009). Evaluación de las Prácticas Pesqueras This initiative has been a collaborative effort between multiple in‐ en Pesquerías de Pequeña Escala del Golfo de Fonseca, Honduras. stitutions, including government departments, non‐government Recomendaciones para el Manejo, 39. organisations and local communities. The system supports the aim Box, S. J., & Canty, S. W. J. (2011). The Long and Short Term Economic Drivers of Overexploitation in Honduran Fisheries Due to provide managers with up‐to‐date, comprehensive fisheries data to Their Dependence on Export Markets (p. 9). Presented at the on an ongoing basis, which can be used to develop local, regional Proceedings 63rd Gulf and Caribbean Fisheries Institute, San Juan, and national fisheries management plans, and directly include and Puerto Rico. link principle stakeholders (i.e. local artisanal fishers and fish buyers) CITES (Convention on International Trade in Endangered Species). 2003. Review of significant trade in specimens of Appendix II species within the data collection and data use process. With the advent of (Resolution Conf. 12.8 and Decision 12.75). CITES. AC19 Doc.8.3. these digital platforms that can be used directly by fishing communi‐ Davies, R. W. D., Cripps, S. J., Nickson, A., & Porter, G. (2009). Defining ties, the onus is now on those involved in fisheries management and and estimating global marine fisheries bycatch. Marine Policy, 33, governance to help scale‐up the adoption of these systems to trans‐ 661–672. Espinoza, E. (2007). Mejoramiento de los mercados internos de produc‐ form fisheries data collection to support data for decision making tos pesqueros en América Latina, Caso de Honduras (No. FAO – TCP/ around the status and use of marine fisheries. Finally, these fisheries RLA/311 Project). Tegucigalpa, Honduras. management and governance participants also need to ensure that FAO (2002). Resumen Informativo sobre pesca por países: Honduras (FID/ these new data are seamlessly incorporated in all national and inter‐ CP/HND). Roma, UN: FAO. FAO. (2016). The state of world fisheries and aquaculture. Contributing to national (i.e. FAO) data reporting schemes. food security and nutrition for all(p. 200). Roma, UN: FAO. Froese, R., & Pauly, D. (Eds) (2018). FishBase. World Wide Web elec‐ tronic publication. https://www.fishbase.org ACKNOWLEDGMENTS Funes, M., Zylich, K., Divovich, E., Zeller, D., Lindop, A., Pauly, D., & Box, S. (2015). Honduras, a fish exporting country: preliminary re‐ SC and SB were funded by the Summit Foundation; MF received constructed marine catches in the Caribbean Sea and The Gulf of funding from the Consejo Nacional de Investigaciones Científicas Fonseca, 1950 – 2010, (90), 17. y Técnicas (CONICET); KZ, BD, ED, AL, DP and DZ thank the Sea Gobert, B., Berthou, P., Lopez, E., Lespagnol, P., Turcios, M. D. O., Around Us and the Sea Around Us—, scientific initiatives Macabiau, C., & Portillo, P. (2005). Early stages of snapper–grou‐ at the University of British Columbia and the University of Western per exploitation in the Caribbean (Bay Islands, Honduras). Fisheries Research, 73(1), 159–169. Australia, respectively. The Sea Around Us research is supported by Golden, C., Allison, E. H., Cheung, W. W. L., Halpern, B. S., McCauley, D. the Oak Foundation, the Marisla Foundation, the Paul M. Angell J., Smith, M., … Myers, S. S. (2016). Fall in fish catch threatens human Family Foundation, the David and Lucile Packard Foundation. This is health. Nature, 534(7607), 317–320. Smithsonian Marine Station at Ft. Pierce, Florida, Contribution No. Haas, A., Harper, S., Zylich, K., & Zeller, D. (2015). Reconstruction of Nicaragua's fisheries catches: 1950‐2010 (Fisheries Centre Working 1108. The authors would also like to thank the Centro de Estudios Paper No. #2015‐23) (p. 10). Vancouver, Canada: University of British Marinos (CEM) and the Bay Islands Conservation Association (BICA) Columbia. for providing valuable information. Harper, S., Guzman, H. M., Zylich, K., & Zeller, D. (2014). Reconstructing Panama's Total Fisheries Catches from 1950 to 2010: Highlighting Data Deficiencies and Management Needs. Marine Fisheries Review, ORCID 76(1–2), 51–65. Heyman, W. D., & Granados‐Dieseldorff, P. (2012). The voice of the fish‐ Steven Canty https://orcid.org/0000-0001-9927-7736 ermen of the : Improving regional fisheries man‐ agement through fisher participation. Fisheries Research, 125–126, 129–148. Hughes, T. P. (1994). Catastrophes, phase shifts, and large‐scale degrada‐ REFERENCES tion of a caribbean coral reef. Science, 265(5178), 1547–1551. Belhabib, D., Campredon, P., Lazar, N., Sumaila, U. R., Baye, B. C., Kane, INE. (2018). Social and economic stats. Retrieved from http://www.ine. E. A., & Pauly, D. (2016). Best for pleasure, not for business: evaluat‐ gob.gt/np/ ing recreational marine fisheries in West using unconventional Korda, R. C., Hills, J. M., & Gray, T. S. (2008). Fishery decline in Utila: sources of data. Palgrave Communications, 2(1), 15050. https://doi. Disentangling the web of governance. Marine Policy, 32(6), 968–979. org/10.1057/palcomms.2015.50 Kramer, P. A., & Kramer, P. R. (2002). Ecoregional Conservation Belhabib, D., Koutob, V., Sall, A., Lam, V. W. Y., & Pauly, D. (2014). Fisheries Planning for the Mesoamerican Caribbean Reef [Technical catch misreporting and its implications: The case of Senegal. Fisheries Reports]. Retrieved October 17, 2018, from http://eprints.uberibz. Research, 151, 1–11. org/19/ CANTY et al. | 11

Le Manach, F., Chavance, P., Cisneros-Montemayor, A. M., Lindop, A., species from reduction fisheries. Frontiers in Marine Science, 4, 1–10. Padilla, A., Schiller, L., & Pauly, D. (2016). Global catches of large pelagic https://doi.org/10.3389/fmars.2017.00363 fishes, with emphasis on the high . In Global Atlas of Marine Fisheries: Trujillo, P., Cisneros‐Montemayor, A. M., Harper, S., Zylich, K., & Zeller, Impacts and Analysis. Washington, D.C.: Island Press. D. (2012). Reconstruction of Costa Rica's marine fisheries catches, MacKenzie, J., & Stehlik, L. L. (1996). The crustacean and molluscan fish‐ 1950‐2010 (Fisheries Center Working Paper No. #2015‐31) (p. 16). eries of honduras. Marine Fisheries Review, 58, 33–44. Vancouver, Canada: University of British Columbia. Muggeo, V. M. R. (2008). Segmented: An R package to fit regression Zeller, D., Booth, S., Davis, G., & Pauly, D. (2007). Re‐estimation of models with broken‐line relationships. R News, 8, 20–25. small‐scale fishery catches for U.S. flag‐associated island areas in Pauly, D. (1998). Rationale for reconstructing catch time series. EC the western Pacific: The last 50 years. Fishery Bulletin, 105, 266–277. Fisheries Cooperation Bulletin, 11(2), 4–7. Retrieved from http://www.vliz.be/en/imis?refxml:id=128748 Pauly, D. (2006). Rejoinder: Towards consilience in small‐scale fisheries Zeller, D., Graham, R., & Harper, S. (2011). Reconstruction of total ma‐ research. MAST, 4(2), 47–51. rine fisheries catches for Belize, 1950‐2008. In M. L. D. Palomares Pauly, D., & Charles, A. (2015). Counting on small‐scale fisheries. Science, & D. Pauly (Eds.), Too Precious to Drill: The Marine Biodiversity of 347(6219), 242–243. Belize, Vol. 19 (pp. 142–151). Vancouver, Canada: University of Pauly, D., Christensen, V., Guénette, S., Pitcher, T. J., Sumaila, U. R., British Columbia. Walters, C. J., … Zeller, D. (2002, August 8). Towards sustainabil‐ Zeller, D., Harper, S., Zylich, K., & Pauly, D. (2015). Synthesis of underre‐ ity in world fisheries [Special Features]. https://doi.org/10.1038/ ported small‐scale fisheries catch in Pacific island waters. Coral Reefs, nature01017 34(1), 25–39. Pauly, D., & Zeller, D. (2016a). Catch reconstructions reveal that global Zeller, D., Palomares, M. L. D., Tavakolie, A., Ang, M., Belhabib, D., marine fisheries catches are higher than reported and declining. Cheung, W. W. L., … Pauly, D. (2016). Still catching attention: Sea Nature Communications, 7, 10244. Around Us reconstructed global catch data, their spatial expression Pauly, D., & Zeller, D. (2016b). Global Atlas of Marine Fisheries: A Critical and public accessibility. Marine Policy, 70, 145–152. Appraisal of Catches and Ecosystem Impacts. San Salvador: Island Zeller, D., & Pauly, D. (2018). The ‘presentist bias’ in time‐series data: Press. Implications for fisheries science and policy. Marine Policy, 90, 14–19. Perez, A. (2009). Fisheries management at the tri‐national border Zylich, K., Harper, S., Licandeo, R., Vega, R., Zeller, D., & Pauly, D. (2014). between Belize, Guatemala and Honduras. Marine Policy, 33(2), Fishing in Easter Island, a recent history (1950‐2010). Latin American 195–200. Journal of Aquatic Research, 42(4), 845–856. Retrieved from http:// Schiller, L. L. (2014). Tuna be, or not tuna be : Using catch data to observe www.redalyc.org/resumen.oa?xml:id=175032366011 the ecological impacts of commercial tuna fisheries in the Pacific Ocean at varying spatial scales. Vancouver, Canada: University of British Columbia. https://doi.org/10.14288/1.0166022 SUPPORTING INFORMATION Soto, L. (2012). Informe de evaluación de la actividad pesquera en el Golfo de Fonseca. Additional supporting information may be found online in the Sumaila, U. R., Marsden, A. D., Watson, R., & Pauly, D. (2007). A global Supporting Information section at the end of the article. ex‐vessel fish price database: Construction and applications. Journal of Bioeconomics, 9(1), 39–51. Swartz, W., Sumaila, R., & Watson, R. (2013). Global ex‐vessel fish price How to cite this article: Canty S, Funes M, Box S, et al. The database revisited: A new approach for estimating ‘Missing’ prices. Environmental and Resource Economics, 56(4), 467–480. hidden value of artisanal fisheries in Honduras. Fish Manag Tai, T. C., Cashion, T., Lam, V. W. Y., Swartz, W., & Sumaila, U. R. (2017). Ecol. 2019;00:1–11. https://doi.org/10.1111/fme.12346 Ex‐vessel fish price database: Disaggregating prices for low‐priced